Cool-flame combustion

The successful isolation of substantial amounts of O-heterocycles will depend to a considerable extent on the design of the reaction vessel. Thus, for example, high yields of such compounds are produced in the falling cloud reactor under conditions corresponding to cool-flame combustion (10, 11). Comparatively little information is available, however, as to the variation in yield with molecular structure of the initial hydrocarbon. 

In the present work, therefore, a comparative study of the production of O-heterocycles during the cool-flame combustion of three consecutive n-alkanes—viz., n-butane, n-pentane, and n-hexane—was carried out under a wide range of reaction conditions in a static system. The importance of carbon chain length, mixture composition, pressure, temperature, and time of reaction was assessed. In addition, the optimum conditions for the formation of O-heterocycles and the maximum yields of these products were determined. The results are discussed in the light of currently accepted oxidation mechanisms. 

This work has thus shown that considerable quantities of O-hetero-cycles may be formed during the cool-flame combustion of n-alkanes... 

Barusch and Payne (6), in 1951, were successful in stabilizing a cool flame in a straight tube, and used this device to investigate the relationship between the octane number of the fuel and its tendency to form cool flames. Using a similar device Ober-dorfer (30), working in the author s laboratory was able to study the cool flames of the isomeric hexanes. In this manner, n-hexane, 2-methyl pentane, 3-methylpentane, and 2,2-dimethylbutane were readily brought to cool-flame combustion. The fuel-air ratio... 

In addition to the monocarbonyls, olefins appear as intermediate cool-flame combustion products. At first these were estimated simply as ethylene. Later Watters (43) and his associates found it possible, by the application of infrared techniques, to estimate the individual olefins in such mixtures. The data are shown in Table II. 

The analytical data on monocarbonyls, olefins, and oxides of carbon were then used to strike a carbon balance for the purpose of finding what part of the cool-flame combustion products had been accounted for. The balance for n-hexane proved to be nearly quantitative. For the more highly branched isohexanes the carbon balance was much less complete. 

Since the work described above was completed, 23 additional hydrocarbons have been brought to cool-flame combustion in the same apparatus (19). Any paraffin, cycloparaffin, or olefin having a research octane number not to exceed 90 may be brought to cool-flame combustion by such a procedure. Aromatic hydrocarbons for the most part do not fall within the indicated range, but n-butyl benzene yields a cool flame with much smoke and soot formation. 

The effect of temperature is unusual as cool flame combustion reactions show a region of temperatures and pressures in which the rate shows anomalous behaviour the rate decreases as the temperature increases - the negative temperature coefficient effect. 

Recently Halstead et aL [122, 122a] have proposed a model for acetaldehyde cool flame combustion which is basically similar to that described above. Their treatment accounts for the periodicity and the self-quenching which is attributed to a thermal switch in which the decomposition of CH3CO... 

A similar sharp transition has been noted [43] when formaldehyde is the additive. Furthermore, there is an accompanying shift of the cool flame limits to higher pressures and/or temperatures. Formaldehyde also interferes with the cool flame combustion of acetaldehyde in flow systems [7]. Its addition retards the development of the cool flame but promotes the second stage this promotion may be caused by the occurrence of... 

Much of the outstanding chemical investigation of the low-temperature combustion of alkanes was performed on the C5 and alkanes, in the late 1960s and early 1970s, mainly by Cullis and co-workers [137, 189-193] and Fish and co-workers [99, 194-199]. The quality and extent of the chemical analyses in these studies has rarely been equalled in subsequent work, and the data obtained provide very strong evidence, not only for the importance of alkylperoxy radical formation and isomerization in the low-temperature chemistry, but also the role of these reactions in the development of cool-flames and two-stage ignitions. However, one constraint on quantitative application is that much of the information was obtained under markedly non-isothermal conditions in the absence of any record of the reactant temperature. Moreover, the effects of convection on the movement of cool-flame combustion waves within an unstirred reaction vessel were not appreciated at the time [52], which casts doubt on some of the mechanistic interpretations of the evolution of multiple cool-flames that were made [195]. 

The sequence of cool flames is one of the indications of the periodicity of chemical processes, particularly characteristic of cool flame combustion. The possibility of periodicity in a chemical process, accounted for by purely kinetic... 

Cool Flames. Under particular conditions of pressure and temperature, incomplete combustion can result in the formation of intermediate products such as CO. As a result of this incomplete combustion, flames can be less exothermic than normal and are referred to as cool flames. An increase in the pressure or temperature of the mixture outside the cool flame can produce normal spontaneous ignition (1). 

It is essential to establish the specific mechanisms that explain the cool flame phenomenon, as well as the hydrocarbon combustion characteristics mentioned earlier. Semenov [14] was the first to propose the general mechanism that formed the basis of later research, which clarified the processes taking place. This mechanism is written as follows ... 

The slow combustion reactions of acetone, methyl ethyl ketone, and diethyl ketone possess most of the features of hydrocarbon oxidation, but their mechanisms are simpler since the confusing effects of olefin formation are unimportant. Specifically, the low temperature combustion of acetone is simpler than that of propane, and the intermediate responsible for degenerate chain branching is methyl hydroperoxide. The Arrhenius parameters for its unimolecular decomposition can be derived by the theory previously developed by Knox. Analytical studies of the slow combustion of methyl ethyl ketone and diethyl ketone show many similarities to that of acetone. The reactions of methyl radicals with oxygen are considered in relation to their thermochemistry. Competition between them provides a simple explanation of the negative temperature coefficient and of cool flames. 

In a discussion of the mechanism it is convenient to consider three combustion regimes, one operating above about 400°C., another below 320°C., and the third between these temperatures, and within which under appropriate conditions cool flames may occur. 

Figure 6 shows the variation of peroxide concentration in methyl ethyl ketone slow combustion, and similar results, but with no peracid formed, have been found for acetone and diethyl ketone. The concentrations of the organic peroxy compounds run parallel to the rate of reaction, but the hydrogen peroxide concentration increases to a steady value. There thus seems little doubt that the degenerate branching intermediates at low temperatures are the alkyl hydroperoxides, and with methyl ethyl ketone, peracetic acid also. The tvfo types of cool flames given by methyl ethyl ketone may arise from the twin branching intermediates (1) observed in its combustion. 

Prescott, Hudson, Foner, and Avery (60) extended the mass-spectrographic technique to the study of composition profiles across a low-pressure, propane-air flame under somewhat lean conditions. The appearance and disappearance of hydrogen, carbon monoxide, ethylene, and acetylene in the flame were demonstrated clearly. The proportion of acetylene was not high. Nonetheless, it is evident that the formation of acetylene is not just a result of pyrolysis of excess hydrocarbon by heat released in combustion of part of the gas. It is a result of reactions which must occur to some extent in all hydrocarbon combustion, but which would not be observable except by special techniques, or under conditions—such as rich flames or cool flames—where the later reactions of acetylene can l>e minimized. 

Formaldehyde is a product of the combustion of all hydrocarbons. Studies of the reactions of formaldehyde are important in leading to a better understanding of the mechanism of hydrocarbon oxidation. Its role in the low temperature region is variable but minor, and depends on the individual hydrocarbon and conditions. In sufficient quantities it appears able to suppress cool flames. In hydrocarbon oxidation above 400° C. formaldehyde is an important intermediate responsible for degenerate-chain branching. 

Motored engine cool flames are discernible with the naked eye only with great difficulty, hence the reliance on detection apparatus for quantitative investigations. The essential features of radiation detection apparatus, such as shown in Figure 1, include a quartz window mounted in the combustion chamber, a photomultiplier radiation detector, and an amplifier-oscilloscope arrangement (107). 

The nature of the radiation processes is not fully understood. Ball (10,11), with the aid of a stroboscopic shutter, visually observed cool flames as actual flame fronts moving across the combustion chamber of a motored engine. This was later confirmed by Getz (53). The source of cool flame emission in tube experiments has been attributed to excited formaldehyde by Emeleus (51) and Gaydon (52). Cool flame spectra in engines obtained by Levedahl and Broida (70) and Downs, Street, and Wheeler (35) were reported to be due to excited formaldehyde. The nature of the blue flame spectra has not been fully explored, although some evidence points to carbon monoxide emission (35). 

In the above example of the combustion of carbon monoxide the time found experimentally in which the pressure reaches a maximum is 0.4 sec the combustion time of a single element according to an estimate based on the theory of flame propagation [11, 12], is less than 0.001 sec. The loss of heat in 0.4 sec is considerable the increase in pressure takes place so slowly that the state of the gas does not change adiabatically upon compression, and despite the compression each element cools after combustion. However, in 0.001 sec the loss of heat is negligibly small and each element in burning does attain the temperature Tp. 

Hydrocarbon combustions are highly complex with very many reactions participating. Nonetheless, a simplified mechanism can be written containing all the essential features to explain cool flame behaviour. 

Since a wide variety of products are formed in combustion reactions, many more reactions should be included in the sequence. However, the nine reactions quoted in the simple mechanism are all that is needed to explain cool flame behaviour. This is discussed in the following worked problem. 

---